Semantic 3D Reconstruction of Furnished Interiors using Laser Scanning

4 Terrestrial Laser Scanning (TLS) technology is increasingly used for the generation of accurate 5 3D models of objects and scenes. But, converting the acquired 3D point cloud data into a repre6 sentative, semantic 3D model of the scene requires advanced processing and skills. This research 7 field is challenging, particularly when considering inhabited, furnished environments that are char8 acterised by clutter and occlusions. This paper presents a TLS data processing pipeline aimed at 9 producing semantic 3D models of furnished office and home interiors. The structure of rooms 10 (floor, ceiling, and walls with window and door openings) is created using Boundary Representa11 tion (B-Rep) models, that not only encode the geometry of those elements, but also their connec12 tivity. Windows and doors are recognized and modelled using a novel method based on molding 13 detection. For the furniture, the approach uniquely integrates smart technology (RFID) that is in14 creasingly used for Facilities Management (FM). RFID tags attached to furniture as sensed at the 15 same time as laser scanning is conducted. The collected IDs are used to retrieve discriminatory 16 geometric information about those objects from the building’s FM database, that this information 17 is used to support their recognition and modeling in the point cloud data. The manuscript partic18 ularly reports results for the recognition and modeling of chairs, tables and wardrobes (and other 19 similar objects like chest of drawers). Extended experimentation of the method has been carried 20 out in real scenarios yielding encouraging results. 21